Relativistic jets from evolving accretion discs

Rob Fender (Southampton / Amsterdam)

Thanks to:

Elena Gallo, Simone Migliari, Tom Maccarone (Amsterdam)Tomaso Belloni (Brera), Stephane Corbel (Saclay), Peter Jonker (CfA), Guy Pooley (MRAO), Ralph Spencer (Jodrell Bank)

Use X-ray binaries to study the inflow:outflowcoupling around black holes

Relation to AGN ?

Timescales linearly proportional to mass

(is this really true ?

microphysics ?

environment ? )The high-energy astrophysics occurs very close to the compact object…which means the companion star is just a distant orbiting fuel tank…

The basics:

1. Hard X-ray states (truncated accretion disc plus strong ‘corona’) make jets

2. Soft X-ray states (‘full’ accretion disc, much weaker corona) do not make jets

3. Transient outbursts – very rapid state changes –correspond to powerful massive ejection events

The life and times of a black hole X-ray binary…

X-ray hardness

Inte

nsit

y /

Eddi

ngto

n

<10-6

~0.01

~0.1

~1.0

softspectrum

hardspectrum

after, typically,10+ years ofrelative peace, accretion rateincreases

Outburst

source getsvery bright and ‘softens’

source remainssoft for some time then fadesaway, returningto hard X-rayspectrum

Quiescence

So what is the relation to jets ?

X-ray hardness

Inte

nsit

y /

Eddi

ngto

n

<10-6

~0.01

~0.1

~1.0

Steady jets(~1014 cm)

Quiescencesoftspectrum

hardspectrum

From quiescence to the brightest hard X-ray states, there seems to be a steady, powerful, jet.

So what is the relation to jets ?

In steady soft X-ray states there appears to be no jet produced

~1.0

~0.1In

tens

ity

/ Ed

ding

ton

~0.01

<10-6 Quiescencesoftspectrum

hardspectrum

hardnessX-ray

Lradio a LX0.7 in low/hard state (implies jet-dominated states)Ra

dio

luminos

ity

X-ray luminosity

(Corbel et al. 2003; Gallo, Fender & Pooley 2003)

Why this relation ? MHD jets and B scale height ?

"Corona"(base of jet?)

Accretion disc

Low/hard X-ray state

X-ray spectrum dominated bynon-accretion-disc emission

Jets always present

X-ray spectrum dominated by accretion disc emission

B field

B field

High/soft X-ray stateJets never present

X-ray hardness

softspectrum

hardspectrum

Outburst

Early on in a major outburst, a large-scale, very powerful jet is produced…

No good explanation for this… :

~1.0

~0.1In

tens

ity

/ Ed

ding

ton

~0.01

Discrete ejections(up to parsec scales)

<10-6

(Corbel & Fender 2001; Homan, Buxton, Markoffet al. submitted)

Radio flare corresponds to transient production of a relativistic (Γ>2) jet, while core radio emission (central compact jet) is switched off(Gallo et al. 2004)

GX 339-4: a case study of a black hole with varying accretion rate(~7 solar mass black hole, with a faint, <1 solar mass, companion)

Observations demonstrate in hard state that entire spectrum longwards of near-infrared is dominated by the self-absorbed jet (and since this is radiatively inefficient, total power is large)

Then, at state transition, powerful transient jet is formed and steady (core) jet is killed off…

fading

GX339-4

The transient relativistic jet is formed at the soft peak…What occurs at this soft peak ??

Fits to X-ray spectra (ha ha) invariably indicate that this corresponds to the point of minimum inner disc radius (which is often sustained for a period of >100 days)radio flare

The optically thin radio flares occur around the time that the optically thick accretion disc reaches its innermost radius

… but why ? ? ?

(e.g. Zdziarski et al. 2004)

jet speed variations with Lx or state?

low/hard state limit

Cir X-1

Sco X-1

V4641 Sgr

1915, 1655, 1748,1550, GX 339-4…

GX 339-4 is inboth samples (onesource can change Γ)

Jet velocity increases with increasing LX : but not clear if velocity function is a ‘step’ or smooth… (Fender, Belloni & Gallo 2004)

Disc moves in, Γjet increases Internal shock

Γ>2 Γ<2

1. ‘Hard’ X-ray state: steady, Γ<2, self-absorbed jet

2. Disc moves in, ‘softening’ the X-ray spectrum. A transient, high-Γ flow collides with the slower pre-existing jet

internal shocks.

‘Corona’, not disc, ejected(i.e. GRS 1915+105)

Obviously you only get the shock when δΓ>0 (so only one flare per ‘cycle’)τ~1

(almost like an ‘external’ shock with steady jet as ambient medium)(see also Kaiser, Sunyaev & Spruit 2000; Vadawale et al. 2004; Turler et al. 2004)

Towards a unified model…

Faint, hard source have steady, Γ~1 jets

More powerful, hard sources have more powerful, steady jets…

As source softens, jet velocity increases abruptly, causing internal shock in jet

Subsequently, soft states show no jet

Only crossing the ‘jet line’ from hard to soft makes an outburst !!

Some sources do this more than once…1915?

1859, 1550?

Crossing from soft to hard (e.g. quiescence) there is no shock

Generic: Fender, Belloni & Gallo (2004)1915: Fender & Belloni (2004)

Hard state jet (steady state)(VLBA)

Cygnus X-1: three phases of jet from black hole to ISM

zoom out x 10: transient jet at state change(internal shock over several hours)

(MERLIN)

zoom out x 50 000: jet-ISM interaction (external shock over 106 years) (WSRT)

(Stirling et al. 2001; Stirling, Fender in prep; Gallo et al. in prep)

A single function may fit steady/transient jetsOr, there may be a step up to a more powerful mode (connection to black hole spin ?)

JET POWER ?

steady jet – no jet – transient jet

(Fender, Gallo & Jonker2003

Malzac, Merloni & Fabian 2004

Fender, Belloni & Gallo 2004)

JDAF

JDAF

Ljet α LX0.5 Jet dominated states

(for MMF04 estimate, all hard states are jet-dominated)

Inte

nsit

y /

Eddi

ngto

n

<10-6

~0.01

~0.1

~1.0

X-ray hardness

softspectrum

hardspectrum

Quiescence<10-9

This is the regime of the Lradio α correlation

This is where the jet shuts off

Could XRB behaviour apply to AGN ?... … and do transitions occur at same LX/LEdd ?(surely they must?) LX

0.7

Merloni, Heinz & di Matteo (2003), see also Falcke, Koerding & Markoff (2004)

Extending the LR:Lx correlation to AGN – add a mass term?

where is effect ofBH spin ?

Schwarzschildblack holes ?

Kerr blac

k holes ?

‘Quenching’ of radio emission in the same (Eddington fraction) luminosity range as the XRB soft state… ?

XRBs

AGN

Quantitative comparisons can be made between XRBsand AGN

steady jet – no jet – transient jet

Maccarone, Gallo &Fender (2003)

Is the fact that the central object is a black hole important … ?Neutron stars can act as a control sample.

Neutron stars behave qualitatively the same but quantitatively differently to black holes

(same patterns but are less ‘radio loud’ for a given LX)

(Migliari et al. 2003, 04, 05)

Conclusions and speculations…

1. Accretion flow state and behaviour of jet strongly coupled. Jets do not like thin discs. Every ‘ADAF’case has a jet.

2. There is a common pattern of behaviour in outbursts, which might be explained by an internal shock as jet velocity increases just before it is ‘quenched’ by disc…

3. These patterns may actually apply qualitatively and (!) quantitatively to AGN

4. Acting as a control sample, neutron stars show that ‘type’ of jet relates to disc only, not nature of accretor… BUT black holes seem to be more ‘radio loud’… The end